Inorganic Chemistty, Val. 12, No. 2,1973 323
OH- as a Reducing Agent for Ru Trimers Table 111. Mutual Atom Polarizabilities? nr.1 = aqr/aQ1 Atom (r) p P
=1 =2 a
1
2
+0.273 +0.166
-0.096 -0.068
HMO calculations; CYN= ap
4
5
-0.006 -0.003
-0.023 -0.004
3 -0.023 -0.010
+ pp; homomorphic.
by an addition reaction, which can be understood similarly, and a preliminary account of this work has appeared.lg Conjugative,” solvent, and reagent effects are also important and will be considered later. Registry No. MeLi, 917-54-4; N4P4F8,14700-00-6;
MeN4P4F7,371 10-97-7; 1,1-Me2N4P4F6,29021-59-8; trans-1,5-Me2N4P4F6,371 10-98-8;Me2N4P4F6,3716425-3; 1,1,3-Me3N4P4FS,371 10-93-3;1 , l ,5-Me3N4P4F5, 371 10-94-4; 1,l ,5,5-Me4N4P4F4,29 144-50-1;Me8N4P4, 4299-49-4. Acknowledgments. We thank the National Research Council of Canada for financial support, the University of British Columbia for a University Graduate Fellowship (to T. N. R.), and Dr. L. D. Hall and Mr. R. Burton for the heteronuclear decoupling experiments. (19) N. L. Paddock, T. N. Rpnganathan, and S . M. Todd, Can. J. Chem., 49, 164 (1971). (20) T. Chivers and N. L. Paddock, Inorg. Chem., 11, 848 (1972).
Contribution from the Department of Chemistry, Georgetown University, Washington, D. C. 20007
Hydroxide Ion as a Reducing Agent for Cations Containing Three Ruthenium Atoms in Nonintegral Oxidation States JOSEPH E. EARLEY* and TERENCE FEALEY
Received June 9, 1972
I
Hydroxide ion is oxidized by Ru(NH,),ORu(NH,),ORu(NH,),7c in a second-order reaction for which k is 1.9 X lo4 sec-’ M-’ (at 25“ in buffered media of ionic strength 0.25 M )and AH* is 19 kcal/mol. An isosbestic point exists throughout the reaction but chemical evidence shows than an intermediate is involved. The corresponding oxidant which has two ethylenediamine ligands replacing the four ammonias on the central ruthenium atom i s reduced faster but with a similar A*. The spectra of the four ruthenium trimers involved in these reactions are consistent with a molecular orbital scheme. All the data are consistent with a mechanism involving rate-determining attack of OH- on the central ruthenium atom.
There have been reports’,’ of a curious reaction by which OH- is rapidly oxidized by (NH3)5RuORu(NH3)40Ru(NH3)57+ (“ruthenium brown,” hereafter I) with the production of the corresponding 6f cation (“ruthenium red,” hereafter 11). Previous workers’ suggested that the OH radical was the oxidation product of OH-, but this seems unlikely since the I + I1 potential is only 0.75 V us. the normal hydrogen electrode, which appears insufficient3 to produce OH. Strong oxidants such as MnO, normally react with OH- only quite slowly,4 but the ruthenium trimer I reacts rapidly with OH-. We have investigated this system to determine why such a mild oxidant oxidizes OH- so readily. We recently described5 the synthesis of an analog of the 6+ ion with two ethylenediamine (en) groups replacing the NH3 groups on the central ruthenium atom and also an X-ray structure determination of this ion. We now report spectra of four interrelated compounds (en and NH3, 6+ and 7f). The remarkable reactivity of I with OH- can be rationalized on the basis of a molecular orbital scheme based on these spectra. Experimental Section LiClO, was prepared from Li,CO, and HClO,, digested in so(1) J . M. Fletcher, B. F. Greenfield, C. J. Hardy, D. Scargill, and J . L. Moorhead, J. Chem. Soc., 2000 (1960). (2) J . E. Earley and T. Fealey, Chem. Commun., 331 (1971). (3) P. George and J. S. Griffith, Enzymes, 1, 347 (1959). (4) R. Veprek-Siska, V. Ettle, and A. Regner, Collect. Czech.
Chem. Commun., 31, 1237 (1966). (5) P. M. Smith, T. Fealey, J . E. Earley, and J. V. Silverton, Inorg. Chem., 10, 1943 (1971).
lution overnight at 40-50” to remove SO,, and recrystallized three times from triply distilled water. The chloride salt of “ruthenium red” (11) was prepared as follows. RuCl, .3H,O (5 g) was dissolved ip 25 ml of 0.25 M HCI. Absolute alcohol (5 ml) and ascorbic acid (0.1 g) were added and the brown solution was refluxed at 85’ for 3 hr and then concentrated to 5 ml under reduced pressure. Concentrated NH, (20 ml) was added a few milliliters at a time and the mixture was held at 85-90’ for 1 hr while streams of air and ammonia gas were passed through the solution. Concentrated ammonia solution was added as necessary to maintain the volume of the reacting solution at approximately 20 ml. The reaction mixture was centrifuged while hot, and the supernatant liquid was then cooled in an ice bath. A brown powder was collected by filtration, recrystallized from 0.1 M ammonia, and washed with ethanol and ether (yield 2 n). Anal. Calcd for Ru-0,(NH,),C16.3H,0: Ru, 36.10; N, 23.33; C1, 25.31. Found:6 Ru,36.12; N, 23.10; C1, 27.58. In order to prepare the chloride salt of ruthenium brown (I), the chloride of I1 (1.0 g) was dissolved in water at 40” and 2 M HC1 was added until the pH of a cooled aliquot was 1.O. A stream of air was then passed through the warm solution for several hours. The mixture was cooled and a brown powder was collected by filtration and washed with 0.1 M HC1 and ether (yield 0.8 g). Anal. Calcd for RU,O,(NH,)~,C~,~~H,O~HC~: Ru, 33.82; N, 21.90; C1, 31.69. Found: Ru, 33.48; N, 21.90; C1, 31.18. Needle-shaped crystals were grown by rapid evaporation of solutions of the chloride of I. The ethylenediamine (en) analog of 11, namely, Ru,O,(NH,),,(en),C16 .H,O (hereafter II’), was prepared as previously described. In order tQprepare the oxidized analog (hereafter I’), 50 mg of I1 chloride was dissolved in a minimum quantity of water. The solution was acidified and 2 drops of chlorine water (-10 mM) were added. The color of the solution rapidly changed from red to brown. A few drops of a saturated solution of NaCl in 0.01 M HC1 (6) P. M. Smith, Thesis, Georgetown University, 1971.
324 Inorganic Chemistry, Vol. 12, No. 2, 1973
Joseph E. Earley and Terence Fealey
Table I. Kinetic Data" for Reaction 1 in Buffered LiC10, Media (u = 0.25 M , T = 25.0") PH 6.60 6.60C 6.85C 7.30d 7.36 7.42 7.80 7.95 8.10
1O2kopsd, i0-4k,b secM - ' sec'' 0.06 0.11 0.41 0.48 0.65 0.99 1.28 1.68
1.6 1.45 2.0 2.1 2.5 1.65 1.55 1.4
pH 8.17 8.35e*f 8.65e 8.70C,b 8.85e 9.14f
lo'kopsd, sec-
i04k,b M-' sec-'
3.10 5.36 11.3 11.1 14.4 23.0
2.05 2.2 2.5 2.2 2.05 1.65 1.9 + 0.3g
Av
a Duplicates in Trisma base buffer (0.5 M ) unless otherwise noted. [ I ] = 2.25 X 1O+M. k = kobsd/[OH-]. Imidazole buffer 0.05 M. Single run. e Triplicate. f Borate buffer 0.05 M . g Weighted average and weighted average deviation.
was added and caused a brown microcrystalline precipitate to form. The solid was collected by filtration and washed with 0.01 M HC1 and ether and air-dried. The solid was recrystallized from 0.01 M HC1. Anal. Calcd for Ru30,(en),(NH3),,C1,~5H,0:Ru, 31.79; N, 20.55;C1, 26.05;C,5.03;H,5.87. Found: Ru, 32.9;N, 19.69; C1, 25.86;C,5.41;H,5.84. Imidazole was twice recrystallized from benzene (with charcoal) and dried under vacuum for 2 days. Tris(hydroxymethy1)aminomethane (hereafter Trisma base) and boric acid were reagent grade and used without further purification. Buffer concentrations are expressed as the sum of concentrations of acidic and basic components. Rate measurements were made using Gilford and DurrumGibson kinetic spectrophotometers with temperature control to 0.05 and 0.2", respectively. Kinetic runs were initiated by syringe addition of buffer to an equal volume of acidic substrate solution. Both solutions had previously been deaerated by bubbling with argon. The pH of the reaction was measured after reaction was complete. Potentials were measured vs. the saturated calomel (NaC1) electrode (sce) or saturated mercuric sulfate electrode and are reported Y S . the sce. During the I-OH- reaction, the spectrum characteristic of I disappeared and was replaced by that characteristic of 11. A single, well-defined isosbestic point at 492 nm was observed. Reactions were followed at 532 nm, the wavelength of maximum absorption of 11. The reaction followed first-order kinetics (within 3%) for 3 or more half-times when the initial concentration of I was greater M . Rate plots for those runs involving [ I ] less than than lo-* M showed upward deviations after 1 half-time suggesting some tendency toward a higher order dependence on substrate concentration at extremely low initial concentration of I. For some experiments the substrate (I) solution was purified by chromatography. A 20 X 2.5 cm column of Bio Rex 70 (a weak carboxylic acid resin) in the H+ form was loaded with I in 1 m M HClO,. The column was then washed with 5 1. of mM HClO, and then with a solution 4 m M in HC10, and 4 m M in LiClO,, which eluted some I1 but did not elute I. I was eluted from the column using a solution -9 m M in HC10, and 4 m M in LiClO,. The concentration of I in the eluted solution was determined spectrophotometrically on the basis of known extinction coefficients. Quantitative formation of I was brought about by adding chlorine water to a solution of I1 which had been formed by adding base t o a solution of I and then, after argon deaeration, reacidifying. This cyclical procedure could be repeated up to ten times (under argon) with the same solution and resulted in only minor decomposition of the trimeric ions.
Results Both at the dme and at the rotating platinum electrode, waves corresponding to the I + I1 and I1 + I changes appeared at +0.49 f 0.03 V vs. the sce. An additional cathodic wave for I1 appeared in the neighborhood of -0.6 V. This wave is assigned to further reduction of 11. The I + I1 redox potential was independent of pH between 2 and 10.
Stoichiometry At the dme, O2 is reduced to H 2 0 2 near 0 V and the latter is further reduced to H 2 0 near - 1.O V. Polarographic experiments designed to detect H 2 0 2 (at -1.0 V) were in-
Table 11. Variation of k with Substrate and Ru3+Concentrationsa
X I
Ib
1061xi . M
k.M-' sec-'
4.0 2.0 0.4 (0)
2.64 1.86 1.64 (1.4)
3.1 1.54 0.74
(0) Ru 3+
0.5C 1.oc
2.68 2.06 1.8 (1.7) 2.4 4.3
a Conditions as in Table I. b Purified by chromatography. = 2 x 10-5 M.
[I]
conclusive, due, in part at least, to high background currents at that potential. Addition of OH- to solutions of I causes an increase in polarographic current near -0.1 V, and this increase can be destroyed by deaeration with nitrogen, s u g gesting that the increase in current is caused by 02. In one series of nine experiments involving a suspension of total composition 0.3 mM in I, an initial current enhancement at -0.1 V corresponding to 60 f 20% of the O2 to be expected on the basis of eq 1 was observed. The product solution is I t OH- -.+ I1 +
t 1/2H20
(1)
identical with a solution prepared from recrystallized I1 in all properties examined except one: if the product solution is reacidified within a few minutes after its formation, part (-20%) of I1 is reconverted to I in a reaction which proceeds at a rapid but measurable rate.7 I1 is oxidized by O2 or H 2 0 2 at rates that are several orders of magnitude less. This difference indicates that there is an intermediate of some stability formed in this reaction, but the existence of the isosbestic point shows that this intermediate has the same chromophore as 11. Kinetics As shown in Table I, observed first-order rate constants depend on [OH-]. The second-order rate constant (k = kobsd/[OH-] ) was constant within the precision of the measurements. Experiments in which C1- or Br- (up to 0.1 M ) were substituted for some of the ClO, (1 5 experiments) or in which buffer concentrations were varied in the range 0.01-0.70M(10 experiments) all gave values of k ' within the range shown in Table I. Values of k in the same range were obtained in the presence of Fe(C104)2, Fe(C104)3, AgBF4, or C6H5NH3C1at concentrations of the order of 0.02 mM, but R U ( H , O ) ~ ( B F ~showed )~ a catalytic effect, approximately doubling the rate when present in 0.001 mM concentration. As shown in Table 11, the observed rate has a small dependence on the initial concentration of substrate and this dependence is reduced, but not entirely eliminated, by chromatographic treatment of the substrate. This behavior and the analytical data are consistent with the presence of a trace (
